growthrates (version 0.8.4)

all_growthmodels: Fit Nonlinear Growth Models to Data Frame

Description

Determine maximum growth rates by nonlinear fits for a series of experiments.

Usage

all_growthmodels(...)
# S3 method for formula
all_growthmodels(
  formula,
  data,
  p,
  lower = -Inf,
  upper = Inf,
  which = names(p),
  FUN = NULL,
  method = "Marq",
  transform = c("none", "log"),
  ...,
  subset = NULL,
  ncores = detectCores(logical = FALSE)
)
# S3 method for `function`
all_growthmodels(
  FUN,
  p,
  data,
  grouping = NULL,
  time = "time",
  y = "value",
  lower = -Inf,
  upper = Inf,
  which = names(p),
  method = "Marq",
  transform = c("none", "log"),
  ...,
  ncores = detectCores(logical = FALSE)
)

Value

object containing the parameters of all fits.

Arguments

...

generic parameters, including parameters passed to the optimizer.

formula

model formula specifying dependent, independent and grouping variables in the form: dependent ~ independent | group1 + group2 + ....

data

data frame of observational data.

p

named vector of start parameters and initial values of the growth model.

lower

lower bound of the parameter vector.

upper

upper bound of the parameter vector.

which

vector of parameter names that are to be fitted.

FUN

function of growth model to be fitted.

method

character vector specifying the optimization algorithm.

transform

fit model to non-transformed or log-transformed data.

subset

a specification of the rows to be used: defaults to all rows.

ncores

number of CPU cores used for parallel computation. The number of real cores is detected automatically by default, but fort debugging purposes it could be wise to set ncores = 1. Usage of logical (hyperthreading) cores does not speed up computation.

grouping

vector of grouping variables defining subsets in the data frame.

time

character vector with name of independent variable.

y

character vector with name of dependent variable.

See Also

Other fitting functions: all_easylinear(), all_splines(), fit_easylinear(), fit_growthmodel(), fit_spline()

Examples

data(bactgrowth)
splitted.data <- multisplit(value ~ time | strain + conc + replicate,
                 data = bactgrowth)

## show which experiments are in splitted.data
names(splitted.data)

## get table from single experiment
dat <- splitted.data[["D:0:1"]]

fit0 <- fit_spline(dat$time, dat$value)

fit1 <- all_splines(value ~ time | strain + conc + replicate,
                 data = bactgrowth, spar = 0.5)

# \donttest{
## these examples require some CPU power and may take a bit longer

## initial parameters
p <- c(coef(fit0), K = max(dat$value))

## avoid negative parameters
lower = c(y0 = 0, mumax = 0, K = 0)

## fit all models
fit2 <- all_growthmodels(value ~ time | strain + conc + replicate,
          data = bactgrowth, FUN=grow_logistic,
          p = p, lower = lower, ncores = 2)

results1 <- results(fit1)
results2 <- results(fit2)
plot(results1$mumax, results2$mumax, xlab="smooth splines", ylab="logistic")

## experimental: nonlinear model as part of the formula

fit3 <- all_growthmodels(
          value ~ grow_logistic(time, parms) | strain + conc + replicate,
          data = bactgrowth, p = p, lower = lower, ncores = 2)

## this allows also to fit to the 'global' data set or any subsets
fit4 <- all_growthmodels(
          value ~ grow_logistic(time, parms),
          data = bactgrowth, p = p, lower = lower, ncores = 1)
plot(fit4)

fit5 <- all_growthmodels(
          value ~ grow_logistic(time, parms) | strain + conc,
          data = bactgrowth, p = p, lower = lower, ncores = 2)
plot(fit5)
# }